Florida is the swing state with the most votes. After going to Bush (republican) in 2008, it went to Obama (democrat) in 2008. Lets look deeper into the financial contribution breakdowns for this election.
Description of columns can be found here
First some preprocessing to download and parse the file
wget ftp://ftp.fec.gov/FEC/Presidential_Map/2008/P00000001/P00000001-FL.zip
unzip P00000001-FL.zipx
mv P00000001-FL.csv 2008-florida-funding.csv
sed -i 's/,$//' 2008-florida-funding.csv # remove trailing comma
Next lets get a summary of the data and data.tables (more efficient)
## [1] 208331 18
## cmte_id cand_id cand_nm contbr_nm contbr_city
## 1: C00431809 P80003478 Huckabee, Mike RAULERSON, JUDY LAKELAND
## 2: C00431809 P80003478 Huckabee, Mike RAULERSON, JUDY LAKELAND
## 3: C00431809 P80003478 Huckabee, Mike RAULERSON, JUDY LAKELAND
## 4: C00431809 P80003478 Huckabee, Mike RAULERSON, JUDY LAKELAND
## 5: C00431809 P80003478 Huckabee, Mike RAULERSON, JUDY LAKELAND
## 6: C00431809 P80003478 Huckabee, Mike REDICK, RICHARD JACKSONVILLE BEACH
## contbr_st contbr_zip contbr_employer contbr_occupation
## 1: FL 33813 CARILLON LAKES REALTY SALES COUNSELOR
## 2: FL 33813 CARILLON LAKES REALTY SALES COUNSELOR
## 3: FL 33813 CARILLON LAKES REALTY SALES COUNSELOR
## 4: FL 33813 CARILLON LAKES REALTY SALES COUNSELOR
## 5: FL 33813 CARILLON LAKES REALTY SALES COUNSELOR
## 6: FL 32250 SELF-EMPLOYED REAL ESTATE
## contb_receipt_amt contb_receipt_dt receipt_desc memo_cd memo_text
## 1: 50 2007-08-03
## 2: 50 2007-08-06
## 3: 50 2007-08-14
## 4: 50 2007-08-29
## 5: 25 2007-09-28
## 6: 250 2007-08-14
## form_tp file_num tran_id election_tp
## 1: SA17A 326973 SA17A.18515 P2008
## 2: SA17A 326973 SA17A.18663 P2008
## 3: SA17A 326973 SA17A.21140 P2008
## 4: SA17A 326973 SA17A.23592 P2008
## 5: SA17A 326973 SA17A.31677 P2008
## 6: SA17A 326973 SA17A.21224 P2008
## cmte_id cand_id cand_nm
## Length:208331 Length:208331 Length:208331
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
##
## contbr_nm contbr_city contbr_st
## Length:208331 Length:208331 Length:208331
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
##
## contbr_zip contbr_employer contbr_occupation
## Length:208331 Length:208331 Length:208331
## Class :character Class :character Class :character
## Mode :character Mode :character Mode :character
##
##
##
## contb_receipt_amt contb_receipt_dt receipt_desc
## Min. :-12800.0 Min. :2005-01-11 Length:208331
## 1st Qu.: 30.0 1st Qu.:2008-01-31 Class :character
## Median : 100.0 Median :2008-05-20 Mode :character
## Mean : 284.2 Mean :2008-04-26
## 3rd Qu.: 250.0 3rd Qu.:2008-09-04
## Max. : 12800.0 Max. :2008-12-31
## memo_cd memo_text form_tp file_num
## Length:208331 Length:208331 Length:208331 Min. :246469
## Class :character Class :character Class :character 1st Qu.:346097
## Mode :character Mode :character Mode :character Median :753761
## Mean :601414
## 3rd Qu.:754317
## Max. :877004
## tran_id election_tp
## Length:208331 Length:208331
## Class :character Class :character
## Mode :character Mode :character
##
##
##
My dataset has ~21k rows and 18 columns. The columns are mostly strings with only one numeric field of interest. This field of interest is donations amount (contb_receipt_amt) and quartiles for donations are between $30 and $250. Quartiles for when donation happened was between January 31st and September 4th of 2008.
Next, I will investigate distribution of the different character columns
## [1] 21
## [1] 20
## [1] 20
##
## C00420224 C00423202 C00430470 C00430512
## Biden, Joseph R Jr 0 0 0 0
## Brownback, Samuel Dale 0 0 0 0
## Clinton, Hillary Rodham 0 0 0 0
## Cox, John H 12 0 0 0
## Dodd, Christopher J 0 0 0 0
## Edwards, John 0 0 0 0
## Gilmore, James S III 0 0 0 0
## Giuliani, Rudolph W 0 0 0 5817
## Gravel, Mike 0 35 0 0
## Huckabee, Mike 0 0 0 0
## Hunter, Duncan 0 0 0 0
## Kucinich, Dennis J 0 0 0 0
## McCain, John S 0 0 31939 0
## Obama, Barack 0 0 0 0
## Paul, Ron 0 0 0 0
## Richardson, Bill 0 0 0 0
## Romney, Mitt 0 0 0 0
## Tancredo, Thomas Gerald 0 0 0 0
## Thompson, Fred Dalton 0 0 0 0
## Thompson, Tommy G 0 0 0 0
##
## C00430694 C00430827 C00430975 C00431171
## Biden, Joseph R Jr 0 0 0 0
## Brownback, Samuel Dale 296 0 0 0
## Clinton, Hillary Rodham 0 0 0 0
## Cox, John H 0 0 0 0
## Dodd, Christopher J 0 0 0 0
## Edwards, John 0 0 0 0
## Gilmore, James S III 0 0 0 0
## Giuliani, Rudolph W 0 0 0 0
## Gravel, Mike 0 0 0 0
## Huckabee, Mike 0 0 0 0
## Hunter, Duncan 0 0 0 0
## Kucinich, Dennis J 0 0 240 0
## McCain, John S 0 0 0 0
## Obama, Barack 0 0 0 0
## Paul, Ron 0 0 0 0
## Richardson, Bill 0 0 0 0
## Romney, Mitt 0 0 0 5792
## Tancredo, Thomas Gerald 0 0 0 0
## Thompson, Fred Dalton 0 0 0 0
## Thompson, Tommy G 0 42 0 0
##
## C00431205 C00431288 C00431379 C00431411
## Biden, Joseph R Jr 0 0 0 0
## Brownback, Samuel Dale 0 0 0 0
## Clinton, Hillary Rodham 0 0 0 0
## Cox, John H 0 0 0 0
## Dodd, Christopher J 0 0 330 0
## Edwards, John 3723 0 0 0
## Gilmore, James S III 0 2 0 0
## Giuliani, Rudolph W 0 0 0 0
## Gravel, Mike 0 0 0 0
## Huckabee, Mike 0 0 0 0
## Hunter, Duncan 0 0 0 84
## Kucinich, Dennis J 0 0 0 0
## McCain, John S 0 0 0 0
## Obama, Barack 0 0 0 0
## Paul, Ron 0 0 0 0
## Richardson, Bill 0 0 0 0
## Romney, Mitt 0 0 0 0
## Tancredo, Thomas Gerald 0 0 0 0
## Thompson, Fred Dalton 0 0 0 0
## Thompson, Tommy G 0 0 0 0
##
## C00431445 C00431569 C00431577 C00431619
## Biden, Joseph R Jr 0 0 0 0
## Brownback, Samuel Dale 0 0 0 0
## Clinton, Hillary Rodham 0 33889 0 0
## Cox, John H 0 0 0 0
## Dodd, Christopher J 0 0 0 0
## Edwards, John 0 0 0 0
## Gilmore, James S III 0 0 0 0
## Giuliani, Rudolph W 0 0 0 0
## Gravel, Mike 0 0 0 0
## Huckabee, Mike 0 0 0 0
## Hunter, Duncan 0 0 0 0
## Kucinich, Dennis J 0 0 0 0
## McCain, John S 0 0 0 0
## Obama, Barack 107907 0 0 0
## Paul, Ron 0 0 0 0
## Richardson, Bill 0 0 1214 0
## Romney, Mitt 0 0 0 0
## Tancredo, Thomas Gerald 0 0 0 485
## Thompson, Fred Dalton 0 0 0 0
## Thompson, Tommy G 0 0 0 0
##
## C00431809 C00431916 C00432914 C00438507
## Biden, Joseph R Jr 0 606 0 0
## Brownback, Samuel Dale 0 0 0 0
## Clinton, Hillary Rodham 0 0 0 0
## Cox, John H 0 0 0 0
## Dodd, Christopher J 0 0 0 0
## Edwards, John 0 0 0 0
## Gilmore, James S III 0 0 0 0
## Giuliani, Rudolph W 0 0 0 0
## Gravel, Mike 0 0 0 0
## Huckabee, Mike 2699 0 0 0
## Hunter, Duncan 0 0 0 0
## Kucinich, Dennis J 0 0 0 0
## McCain, John S 0 0 0 0
## Obama, Barack 0 0 0 0
## Paul, Ron 0 0 5851 0
## Richardson, Bill 0 0 0 0
## Romney, Mitt 0 0 0 0
## Tancredo, Thomas Gerald 0 0 0 0
## Thompson, Fred Dalton 0 0 0 2266
## Thompson, Tommy G 0 0 0 0
##
## C00446104
## Biden, Joseph R Jr 0
## Brownback, Samuel Dale 0
## Clinton, Hillary Rodham 0
## Cox, John H 0
## Dodd, Christopher J 0
## Edwards, John 0
## Gilmore, James S III 0
## Giuliani, Rudolph W 0
## Gravel, Mike 0
## Huckabee, Mike 0
## Hunter, Duncan 0
## Kucinich, Dennis J 0
## McCain, John S 5102
## Obama, Barack 0
## Paul, Ron 0
## Richardson, Bill 0
## Romney, Mitt 0
## Tancredo, Thomas Gerald 0
## Thompson, Fred Dalton 0
## Thompson, Tommy G 0
McCain has two committee ids
##
## Obama, Barack McCain, John S Clinton, Hillary Rodham
## 107907 37041 33889
## Paul, Ron Giuliani, Rudolph W
## 5851 5817
##
## MIAMI TAMPA ORLANDO JACKSONVILLE
## 15414 8351 7342 6393
## BOCA RATON SARASOTA NAPLES TALLAHASSEE
## 6262 6098 6039 6008
## FORT LAUDERDALE MIAMI BEACH GAINESVILLE CORAL GABLES
## 4523 4190 3812 3702
##
## FL
## 208331
##
## 331354726 33139 33133 33480 333212562 33156 320433443
## 398 246 232 194 185 156 152
## 33131 33134 329513361 33629 33432
## 150 147 141 140 121
##
## 33139 33133 33156 33480 33143 33134 32963 32312 32789 33140 33432 33176
## 2453 2336 2154 2124 1711 1697 1465 1464 1454 1248 1208 1202
Cleaned up zip codes, otherwise geography data looks as expected
##
## NOT EMPLOYED
## 41845
## SELF EMPLOYED
## 18968
## RETIRED
## 16844
##
## 12148
## SELF-EMPLOYED
## 6255
## INFORMATION REQUESTED
## 4940
## INFORMATION REQUESTED PER BEST EFFORTS
## 3865
## NONE
## 1930
## HOMEMAKER
## 1476
## SELF
## 1462
## UNIVERSITY OF FLORIDA
## 939
## UNIVERSITY OF MIAMI
## 784
## STATE OF FLORIDA
## 740
## UNEMPLOYED
## 540
## N/A
## 441
##
## NOT EMPLOYED SELF EMPLOYED
## 43775 26685
## RETIRED
## 21394 16844
## HOMEMAKER UNIVERSITY OF FLORIDA
## 1476 939
## UNIVERSITY OF MIAMI STATE OF FLORIDA
## 784 740
## UNEMPLOYED FLORIDA STATE UNIVERSITY
## 540 379
## FLORIDA INTERNATIONAL UNIVERSITY REQUESTED
## 333 300
##
## RETIRED
## 56324
## ATTORNEY
## 10401
## NOT EMPLOYED
## 6633
## HOMEMAKER
## 5892
##
## 5635
## INFORMATION REQUESTED
## 4454
## PHYSICIAN
## 4038
## INFORMATION REQUESTED PER BEST EFFORTS
## 3824
## CONSULTANT
## 2785
## TEACHER
## 2408
## PROFESSOR
## 2401
## PRESIDENT
## 2134
Occupation and employer data is quite messy, a lot of people did not answer. I used blank (“”) to indicate no response. I cleaned it up a bit based on what the top values were. There seems to be a lot of “President” (occupation) who donate. Most of the donors seem to fall under “NOT EMPLOYED”, “SELF EMPLOYED”, “BLANK” or “RETIRED”.
##
## SA17A SA18 SB28A
## 190693 14343 3295
##
## G2006 G2008 P2008
## 38 1 60695 147597
## cmte_id cand_id cand_nm contbr_nm contbr_city
## 1: C00431379 P80003387 Dodd, Christopher J THOMAS, DEROY PALM COAST
## contbr_st contbr_zip contbr_employer contbr_occupation
## 1: FL 321371202 RETIRED
## contb_receipt_amt contb_receipt_dt receipt_desc memo_cd memo_text
## 1: 2100 2006-09-18 X
## form_tp file_num tran_id election_tp contbr_zip2
## 1: SA18 860973 A45C7EB04EC694B87839 G2006 32137
Somehow a donation for 2006 general election snuck into this dataset. I removed it. Not sure why some election type values are blank (38)
Add the party affiliation for each of the candidates and shortened Hillary’s name so make plotting labels easier.
The main feature of interest is donation amount (contb_receipt_amt).
Supporting features include candidate name (cand_nm), donation type (election_tp), donor name (contbr_nm), location (city, zip), work (employer, occupation), date of donation (contb_receipt_dt)
Yes, I added a party field and a proper zip code field.
I cleaned up the zip codes and tried to clean up occupation and employer fields to little success. I converted date field to date object. Donation amount field is sometimes negative.
Next, lets investigate overall contributions to the candidate and the party
Assuming that within the same location (city/zip/employer/occupation) each donor can be uniquely identified by name.
## cand_nm election_tp Total
## 1: Biden, Joseph R Jr P2008 402160.80
## 2: Biden, Joseph R Jr G2008 43200.00
## 3: Brownback, Samuel Dale P2008 94257.97
## 4: Clinton, Hillary P2008 9312563.28
## 5: Clinton, Hillary G2008 303727.94
## 6: Cox, John H P2008 1815.00
## 7: Dodd, Christopher J P2008 351009.10
## 8: Dodd, Christopher J G2008 22300.00
## 9: Edwards, John P2008 1097018.71
## 10: Edwards, John G2008 189944.10
## 11: Edwards, John -250.00
## 12: Gilmore, James S III P2008 1025.00
## 13: Giuliani, Rudolph W P2008 4734594.91
## 14: Giuliani, Rudolph W G2008 -696779.18
## 15: Gravel, Mike P2008 16595.00
## 16: Huckabee, Mike P2008 1181612.94
## 17: Huckabee, Mike -32167.16
## 18: Hunter, Duncan P2008 22100.00
## 19: Kucinich, Dennis J P2008 36977.77
## 20: Kucinich, Dennis J 700.00
## 21: McCain, John S P2008 12735326.32
## 22: McCain, John S G2008 2218187.95
## 23: Obama, Barack P2008 10919619.36
## 24: Obama, Barack G2008 9407917.07
## 25: Obama, Barack 300.00
## 26: Paul, Ron P2008 1231651.99
## 27: Richardson, Bill P2008 796248.79
## 28: Richardson, Bill G2008 1000.00
## 29: Romney, Mitt P2008 4083366.18
## 30: Romney, Mitt G2008 -305375.00
## 31: Tancredo, Thomas Gerald P2008 49048.00
## 32: Thompson, Fred Dalton P2008 923130.16
## 33: Thompson, Fred Dalton G2008 -4880.00
## 34: Thompson, Tommy G P2008 60200.00
## 35: Thompson, Tommy G -500.00
## cand_nm election_tp Total
Unsurprisingly, the candidates who raised the most money were the ones that ran for president (McCain and Obama). Multiple candidates with negative donations had large donations during primary but dropped out during general election.
## cand_nm Party Total
## 1: McCain, John S R 12735326.32
## 2: Obama, Barack D 10919619.36
## 3: Clinton, Hillary D 9312563.28
## 4: Giuliani, Rudolph W R 4734594.91
## 5: Romney, Mitt R 4083366.18
## 6: Paul, Ron R 1231651.99
## 7: Huckabee, Mike R 1181612.94
## 8: Edwards, John D 1097018.71
## 9: Thompson, Fred Dalton R 923130.16
## 10: Richardson, Bill D 796248.79
## 11: Biden, Joseph R Jr R 402160.80
## 12: Dodd, Christopher J D 351009.10
## 13: Brownback, Samuel Dale R 94257.97
## 14: Thompson, Tommy G R 60200.00
## 15: Tancredo, Thomas Gerald R 49048.00
## 16: Kucinich, Dennis J D 36977.77
## 17: Hunter, Duncan R 22100.00
## 18: Gravel, Mike D 16595.00
## 19: Cox, John H R 1815.00
## 20: Gilmore, James S III R 1025.00
There is a large difference in donations from florida during the primary between the McCain and the other Republican candidates.
Focusing on the top fundraisers from the primary, you can see that contributions for the runnerups dried up or went negative (donors withdrawing money) when they did not win the primary. Contributions for the two candidates that won the primaries (Obama and McCain) kept increasing.
Many small donations, some are negative.
## Party Total
## 1: R 26741976
## 2: D 32455671
Overall, democrats raised more money from florida than republicans
Next, lets have a look at which groups donated the most.
## contbr_nm cand_nm contbr_city
## 1: CASTRO, JUANITA C Clinton, Hillary CORAL GABLES
## 2: SCRIBANTE, LYNDA H. MRS. McCain, John S SANIBEL
## 3: BOLANOS, ZOILA GLORIA MS. McCain, John S WESTON
## 4: SMITH, REBECCA J. MS. McCain, John S TAMPA
## 5: BALZOLA, GABRIEL Obama, Barack MIAMI
## 6: SENESE, VICTOR MR. McCain, John S PALM BEACH GARDENS
## 7: SHALLER, NELSON MR. McCain, John S BOCA RATON
## 8: PRESCOTT, EDNA B. MS. McCain, John S STUART
## 9: O'DONNELL, MERRY T MS Obama, Barack JUNO BEACH
## 10: KNAPP, AMY Obama, Barack COCONUT GROVE
## 11: NANCE, JONAS Clinton, Hillary INDIAN HARBOUR BEA
## 12: PEDRAZA, RAUL Obama, Barack MIAMI
## 13: KRONGOLD, M. RONALD MR. McCain, John S MIAMI
## 14: GEO GOUP PAC McCain, John S BOCA RATON
## 15: OSI RESTAURANT PARTNERS INC. PAC McCain, John S TAMPA
## 16: WESTREICH, HELENE Clinton, Hillary BAL HARBOUR
## 17: BURTON, DEMETRA Clinton, Hillary AVENTURA
## 18: WEAVER, DAPHNE Clinton, Hillary MANALAPAN
## 19: BEAN, JOHN Obama, Barack SARASOTA
## 20: SEMBLER, BRENT W. MR. McCain, John S SAINT PETERSBURG
## contbr_employer contbr_occupation
## 1: NOT EMPLOYED RETIRED
## 2: HOMEMAKER HOMEMAKER
## 3: G. & C. DE JESUS INC. PRESIDENT
## 4: THE A.D. MORGAN CORPORATION GENERAL CONTRACTOR
## 5: GB CAPITAL REAL ESTATE DEVELOPMENT/INVESTMENT
## 6: RETIRED RETIRED
## 7: IHS DIALYSIS INC CEO
## 8: HOMEMAKER HOMEMAKER
## 9: NOT EMPLOYED RETIRED
## 10: UNITED HEALTH GROUP EXECUTIVE
## 11: WEICHERT REALTORS SALESMAN
## 12: MAGNO INTERNATIONAL PRESIDENT CORPORATE EXECUTIVE
## 13: SELF EMPLOYED REAL ESTATE
## 14:
## 15:
## 16: NOT EMPLOYED NOT EMPLOYED
## 17: BURTON ASSOC OF SO FLA GOVERNMENT RELATIONS
## 18: NOT EMPLOYED HOMEMAKER
## 19: NOT EMPLOYED RETIRED
## 20: THE SEMBLER COMPANY REAL ESTATE EXECUTIVE
## Total
## 1: 13950.00
## 2: 12800.00
## 3: 12300.00
## 4: 12300.00
## 5: 11500.00
## 6: 11500.00
## 7: 11500.00
## 8: 10900.00
## 9: 10850.00
## 10: 10645.18
## 11: 10200.00
## 12: 10142.50
## 13: 10000.00
## 14: 10000.00
## 15: 10000.00
## 16: 9900.00
## 17: 9200.00
## 18: 9200.00
## 19: 9200.00
## 20: 9200.00
Quite a few large donations supporting McCain. The occupations and employers do not appear to have a pattern.
## contbr_city cand_nm Total
## 1: MIAMI Obama, Barack 2032341.4
## 2: MIAMI Clinton, Hillary 1346698.8
## 3: NAPLES McCain, John S 925217.0
## 4: TAMPA Obama, Barack 891495.0
## 5: MIAMI McCain, John S 868417.3
## 6: MIAMI BEACH Obama, Barack 846736.6
## 7: JACKSONVILLE Obama, Barack 824205.9
## 8: CORAL GABLES Obama, Barack 768285.6
## 9: TALLAHASSEE Obama, Barack 762976.1
## 10: SARASOTA Obama, Barack 711810.0
## 11: JACKSONVILLE McCain, John S 705171.8
## 12: ORLANDO Obama, Barack 675680.5
## 13: TAMPA McCain, John S 668084.5
## 14: BOCA RATON Obama, Barack 663521.8
## 15: BOCA RATON McCain, John S 584809.5
## 16: FORT LAUDERDALE Obama, Barack 583411.0
## 17: NAPLES Obama, Barack 552059.6
## 18: MIAMI BEACH Clinton, Hillary 514105.7
## 19: VERO BEACH McCain, John S 511740.0
## 20: PALM BEACH McCain, John S 489871.0
Miami is a large city and has a lot of donation to many candidates. Naples has one of the highest per-capita income in the US.
## region value
## 1: 17151.99
## 2: 00000 13726.89
## 3: 00001 248.00
## 4: 00003 -40.00
## 5: 03313 1000.00
## 6: 03316 2050.00
## [1] 433 2
I subtracted republican donations in every zip code from democratic donations. Negative (red) zip codes have more republican donation money while positive (blue) zip codes have more democrat donation money.
The area around Tallahasee donated more to democrats while the area around Jacksonville donated more to republicans. The areas around Tampa and Miami are pretty mixed.
Lastly, there are 433 zip codes that do not belong to florida. Next I’ll plot cities instead of zip codes.
## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1358237 72.6 2164898 115.7 2164898 115.7
## Vcells 11854625 90.5 62616268 477.8 97817987 746.3
The candidates raise a lot of money in the area around Miami. Gainsville made top 20 for only Obama and Ron Paul.
During the general election, the top cities that donated mostly overlapped. However, most cities in florida donated more to Obama than McCain.
## contbr_employer cand_nm Total
## 1: NOT EMPLOYED Obama, Barack 5231425.9
## 2: RETIRED McCain, John S 4359725.0
## 3: SELF EMPLOYED Obama, Barack 3563870.5
## 4: NOT EMPLOYED Clinton, Hillary 2781359.0
## 5: SELF EMPLOYED McCain, John S 2028022.5
## 6: SELF EMPLOYED Clinton, Hillary 1613131.3
## 7: Romney, Mitt 1306232.2
## 8: McCain, John S 1216656.0
## 9: HOMEMAKER McCain, John S 1117562.5
## 10: Giuliani, Rudolph W 909353.3
## 11: Obama, Barack 860264.1
## 12: SELF EMPLOYED Giuliani, Rudolph W 635213.3
## 13: SELF EMPLOYED Romney, Mitt 430502.0
## 14: Thompson, Fred Dalton 323546.9
## 15: Clinton, Hillary 259098.2
## 16: Huckabee, Mike 251449.5
## 17: NOT EMPLOYED Paul, Ron 243911.7
## 18: SELF EMPLOYED Edwards, John 234299.2
## 19: NOT EMPLOYED Edwards, John 230850.0
## 20: REQUESTED Huckabee, Mike 202852.7
Sorted by employer, not very useful.
## contbr_occupation cand_nm Total
## 1: RETIRED McCain, John S 4728666.0
## 2: RETIRED Obama, Barack 4007001.2
## 3: ATTORNEY Obama, Barack 2462331.8
## 4: RETIRED Clinton, Hillary 1593216.0
## 5: HOMEMAKER McCain, John S 1213036.5
## 6: ATTORNEY Clinton, Hillary 958995.3
## 7: Obama, Barack 811904.1
## 8: McCain, John S 787892.0
## 9: NOT EMPLOYED Obama, Barack 729532.0
## 10: RETIRED Romney, Mitt 726242.2
## 11: RETIRED Giuliani, Rudolph W 723925.0
## 12: ATTORNEY McCain, John S 719383.7
## 13: HOMEMAKER Clinton, Hillary 655612.1
## 14: HOMEMAKER Obama, Barack 642493.1
## 15: PHYSICIAN Obama, Barack 613105.9
## 16: ATTORNEY Edwards, John 568175.9
## 17: HOMEMAKER Giuliani, Rudolph W 460908.3
## 18: NOT EMPLOYED Clinton, Hillary 447059.0
## 19: HOMEMAKER Romney, Mitt 418004.0
## 20: PHYSICIAN McCain, John S 405667.0
Sorted by occupation, also not very useful.
## contbr_occupation cand_nm Total N
## 1: RESTAURANT BUSINESS McCain, John S 4600.00 1
## 2: PRES. McCain, John S 4600.00 1
## 3: EXECTUTIVE McCain, John S 4600.00 1
## 4: REAL ESTATE & ASSET MANAGEMENT Romney, Mitt 3450.00 2
## 5: CHAIRMAN/PRESIDENT McCain, John S 3450.00 2
## 6: CHAIRMAN Huckabee, Mike 2875.00 4
## 7: AUTOMOBILE DEALER McCain, John S 2425.25 4
## 8: INSURANCE Thompson, Fred Dalton 2300.00 3
## 9: ADMINISTRATIVE ASSISTANT Huckabee, Mike 2300.00 2
## 10: REGIONAL VICE PRESIDENT Huckabee, Mike 2300.00 1
## 11: CHIEF FINANCIAL ADVISOR Richardson, Bill 2300.00 1
## 12: PRESIDENT AND COO Richardson, Bill 2300.00 1
## 13: AUDIO/RECORDING Thompson, Fred Dalton 2300.00 1
## 14: CPA/WEALTH MANAGER Clinton, Hillary 2300.00 2
## 15: GOVERNMENT AFFAIRS CONSULTANT Clinton, Hillary 2300.00 2
## 16: MOVIE PRODUCER Clinton, Hillary 2300.00 2
## 17: BAKER Clinton, Hillary 2300.00 1
## 18: MANAGING PARTNER Richardson, Bill 2300.00 1
## 19: ADMINISTRASION EXECUTIVE Clinton, Hillary 2300.00 2
## 20: AUTHOR Richardson, Bill 2300.00 1
Sorted by average donation per occupation, most occupations only have 1 person. This means that occupations column needs further cleanup.
Boxplots show many outliers. McCain has quite a few donors at the 5k+ level. Most donations are under 3k, I will further investigate this below.
The violin plot shows an interesting difference in distributions. Donations under 3k are often trimodal: a large number of donations around 2.3k, a large number of donations around 1k and a wider group of donors w/a large tail around 100.
Romney and Giuliani both have a larger proportion of donors over 1.5k. Ron Paul and Obama both have more donors who donated less money. For Obama in particular has three large groups of donors who donated under 600.
On average, donation sizes for Guiliani and Romney were larger.
When we aggregate all donations from the same person (assuming one person per name per zip), the average donation increase and difference between candidates decrease. This implies that some donors donate very little but multiple times.
A large number of individuals donated over 10 times to Obama while only 1 person donated for Giuliani. I made the axis the same to make it easier to compare between the different candidates, however, limiting to 75 donations trunates Obama’s dataset.
## cand_nm contbr_nm contbr_city contbr_zip contbr_employer
## 1: Obama, Barack NOGUERA, CARLOTA MIAMI 331354726 NOT EMPLOYED
## 2: Obama, Barack GIARDINA, REBECCA TAMARAC 333212562 NOT EMPLOYED
## 3: Obama, Barack NOGUERA, CARLOTA MIAMI 331354726 NOT EMPLOYED
## contbr_occupation election_tp Party Total_donation N_donations
## 1: HOMEMAKER P2008 D 2300.00 267
## 2: RETIRED P2008 D 2303.13 132
## 3: HOMEMAKER G2008 D 2200.00 116
Suprisingly, 3 people donated over 100 times to Obama’s campaign!
Plotting donations wihtin 2 years of the election date, cumulative donations can be relatively close during the primary but the eventual candidates raise the most money in the end.
At the beginning of January, 2008 the funding totals were
## cand_nm Total
## 1: Clinton, Hillary 6101517
## 2: Giuliani, Rudolph W 5012007
## 3: McCain, John S 2221205
## 4: Obama, Barack 3554218
## 5: Paul, Ron 1006869
## 6: Romney, Mitt 3610918
Suprisingly, the candidates who won the primary did not have the most money raised at this point. Additional donations often stop (plateaus in chart) when each of the candidates drop out (date below). The drops in donations correspond to donors requesting refund.
## Party Total
## 1: R 14116496
## 2: D 12284194
By the end of the election, democrats had raised more money than republicans. However, at the beginning of January 2008 even though Hillary (a democrat) was leading (see numbers above), republicans had raised more money overall. The money raised by republicans was spread over more candidates.
This violin plot shows a lot of information on what the donation size distributions for each of the candidates. I summarize all the donations from an individual (assuming only one donor with same name, location and occupation), remove negative values and plot donation amount from that donor. From this plot, one could make inferences about the net worth of the donors (based on how much they are able to donate).
Removing the restriction of total donation under 3000 magnifies the differences between the candidates’ donor bases. Obama and Ron Paul’s donor base often donate smaller amounts. Romney and Giuliani’s donor base on the other hand donate much more.
## cand_nm Fraction_under_100
## 1: Paul, Ron 0.114002478
## 2: Obama, Barack 0.172848265
## 3: McCain, John S 0.128640904
## 4: Clinton, Hillary 0.140245548
## 5: Romney, Mitt 0.025900566
## 6: Giuliani, Rudolph W 0.005715841
## cand_nm Fraction_over_1000
## 1: Paul, Ron 0.1280463
## 2: Obama, Barack 0.1297157
## 3: McCain, John S 0.2609946
## 4: Clinton, Hillary 0.2236913
## 5: Romney, Mitt 0.4614469
## 6: Giuliani, Rudolph W 0.4730539
The summarized data above quantifies the difference in donor base. Almost half of Romney and Giuliani’s donors donate over $1000 while under 3% donate under $100. In contrast, 17% of Obama’s donors donate under $100.
I chose this plot because it shows the geospatial location of cities that supported each candidate. Most cities supported both candidates with more money going to Obama than McCain. Cities around Miami are the source of most donation donations. The donations plotted are only for general election.
## cand_nm Total city_lookup
## 1: Obama, Barack 924500.53 MIAMI,FL
## 2: McCain, John S 136557.00 NAPLES,FL
## 3: Obama, Barack 401525.44 MIAMI BEACH,FL
## 4: McCain, John S 134075.00 MIAMI,FL
## 5: Obama, Barack 395410.76 TAMPA,FL
## 6: McCain, John S 130270.00 TAMPA,FL
## 7: Obama, Barack 377241.91 CORAL GABLES,FL
## 8: McCain, John S 106302.00 PALM BEACH,FL
## 9: Obama, Barack 324421.72 BOCA RATON,FL
## 10: McCain, John S 76440.50 BOCA RATON,FL
## 11: Obama, Barack 316618.34 TALLAHASSEE,FL
## 12: McCain, John S 74431.85 JACKSONVILLE,FL
## 13: Obama, Barack 313374.15 FORT LAUDERDALE,FL
## 14: McCain, John S 73471.00 VERO BEACH,FL
## 15: Obama, Barack 303758.70 SARASOTA,FL
## 16: McCain, John S 70566.00 ORLANDO,FL
## 17: Obama, Barack 296083.74 JACKSONVILLE,FL
## 18: McCain, John S 61648.57 FORT LAUDERDALE,FL
## 19: Obama, Barack 295288.48 NAPLES,FL
## 20: McCain, John S 41551.00 WINTER PARK,FL
This list of top 10 cities that raised money for Obama and McCain show that while many of the cities overlap, the amount raised differ dramatically.
## cand_nm Total
## 1: Obama, Barack 9407917
## 2: McCain, John S 2218188
## cand_nm Pct city_lookup
## 1: Obama, Barack 9.826835 MIAMI,FL
## 2: McCain, John S 6.156241 NAPLES,FL
## 3: Obama, Barack 4.267953 MIAMI BEACH,FL
## 4: McCain, John S 6.044348 MIAMI,FL
## 5: Obama, Barack 4.202958 TAMPA,FL
## 6: McCain, John S 5.872812 TAMPA,FL
## 7: Obama, Barack 4.009835 CORAL GABLES,FL
## 8: McCain, John S 4.792290 PALM BEACH,FL
## 9: Obama, Barack 3.448391 BOCA RATON,FL
## 10: McCain, John S 3.446079 BOCA RATON,FL
## 11: Obama, Barack 3.365446 TALLAHASSEE,FL
## 12: McCain, John S 3.355525 JACKSONVILLE,FL
## 13: Obama, Barack 3.330962 FORT LAUDERDALE,FL
## 14: McCain, John S 3.312208 VERO BEACH,FL
## 15: Obama, Barack 3.228756 SARASOTA,FL
## 16: McCain, John S 3.181245 ORLANDO,FL
## 17: Obama, Barack 3.147176 JACKSONVILLE,FL
## 18: McCain, John S 2.779231 FORT LAUDERDALE,FL
## 19: Obama, Barack 3.138723 NAPLES,FL
## 20: McCain, John S 1.873196 WINTER PARK,FL
Since the amounts raised by the two candidates differed so much, I also looked into the percent of donations that came from each city. The above table shows that Miami contributed contributed quite a large amount of Obama’s total donations (about 10%) while McCain’s largest city (Naples, FL - one of the wealthiest cities in the US) contributed only about 6% of McCain’s total. About 42% of Obama’s donations for general election came from the top 10 cities while 41% of McCain’s donations for general election came from the top 10 cities.
Some of my early struggles when analyzing this dataset had to do with cleaning up the data and understanding why donations could be negative. I also spent a while trying to figure out the right comparisons, plotting all presidential candidates would make the plots look too complicated so I focused on the top 6 candidates (2 democrats, 4 republicans).
By focusing on the top candidates, the plots were easier to understand. I also spent a while figuring out the libraries to visualize geographically where the zip code boundaries and cities were. When I got them working I felt like it was much more meaningful than just listing zip codes without context.
Some future work that could be done with this dataset is to compare donor zip code to census estimates for population, income and racial breakdown for each zip code. One could also clean up the occupation column and instead use controlled vocabulary instead of free text. Another interesting dataset to overlay would be how many times a candidate visited a city and compare it to rate of donations coming in from that area along with the eventual vote results from election.